Void Code Pattern Detection Tool

MEMA stands for Modified Exponential Moving Average and it is a statistical inquiry tool that functions within SWITCHWARE®.  This tool allows users to set threshold values that trigger a Sentinel™ alert notification once the threshold value is met and is typically used to proactively monitor system void codes.

As an example, a user can identify a void code 92 (timeout) with a threshold value of five.  MEMA will monitor for the common repeated void codes and if the moving average amount for void code 92 becomes greater than five, an alert notification can be sent to the user via Sentinel.  Upon receiving the alert, the user is aware of a recurring issue associated with void code 92 and can research the reason attributable to the void code.

The Moving Average Explained

A moving average is an average for a series of measurements drawn over time, and it is used to remove extreme fluctuations in data. Moving averages involve calculations that smooth fluctuations and extreme outliers to reveal underlying trends.  An exponential moving average uses the same logic except that more weight is given to the most recent data.  In the Modified EMA (MEMA), the length of the averages is “modified” by dividing 1 by the EMA’s length of average. Doing this creates a smoother graphical representation and limits inaccurate spikes in a series of data points. The statistical representation that is presented in the MEMA diagram below shows how this tool is an accurate way for detecting problematic patterns. As the amount of void codes fluctuates over time the MEMA line only fluctuates with the collective average of the entire data set over a moving period, giving emphasis to the most recent data points. Once the MEMA line crosses the threshold point an alert notification can be sent to the appropriate staff member(s).